d8a80f6272
이름만 다른(표기 변형) [[위키링크]]를 대상 문서의 canonical 제목으로 치환해 끊겼던 1,200개 링크를 연결. 제목/파일명 정규화 일치만 적용하고 별칭 매칭은 과병합 위험으로 제외(애매성 가드). 원본은 _link_reconcile_backup/ 에 백업. 도구: Datacollect/scripts/link_reconcile_apply.mjs Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
6.3 KiB
6.3 KiB
id, title, category, status, canonical_id, aliases, duplicate_of, source_trust_level, confidence_score, verification_status, tags, raw_sources, last_reinforced, github_commit, tech_stack
| id | title | category | status | canonical_id | aliases | duplicate_of | source_trust_level | confidence_score | verification_status | tags | raw_sources | last_reinforced | github_commit | tech_stack | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| wiki-2026-0508-server-side-rendering-ssr | Server Side Rendering (SSR) | 10_Wiki/Topics | verified | self |
|
none | A | 0.95 | applied |
|
2026-05-10 | pending |
|
Server Side Rendering (SSR)
매 한 줄
"매 SSR 은 server 에서 HTML 을 생성해 first paint 를 빠르게, SEO 를 가능하게". 2026 기준 React 19 의 streaming SSR + Suspense + Server Components 가 표준. 매 trade-off 는 server compute cost vs CSR-only 의 blank-screen 제거. 매 modern 변형: SSG (build-time), ISR (revalidate), PPR (Partial Prerender).
매 핵심
매 rendering modes
- CSR: 매 client only — slow first paint, no SEO without JS
- SSR: 매 server-render full HTML per request
- SSG: 매 build-time HTML (static)
- ISR: 매 SSG + on-demand or time-based revalidation
- PPR: 매 partial prerender — static shell + dynamic holes (Next.js 15)
매 hydration
- Server HTML 송신 → client JS 가 attach (event handler 연결)
- Streaming SSR: HTML 을 chunk 로 진행 송신
- Selective hydration: visible / interacted 부분 우선
매 응용
- Marketing / blog (SEO + fast paint).
- E-commerce PDP (per-user pricing + SEO).
- Dashboard shells (PPR — static shell + dynamic data).
💻 패턴
Next.js 15 RSC + streaming
// app/page.tsx
import { Suspense } from "react";
import { ProductGrid } from "./product-grid";
export default function Page() {
return (
<main>
<h1>Products</h1>
<Suspense fallback={<div>매 loading…</div>}>
<ProductGrid />
</Suspense>
</main>
);
}
// app/product-grid.tsx (RSC, runs on server)
import { db } from "@/lib/db";
export async function ProductGrid() {
const products = await db.product.findMany({ take: 20 });
return (
<ul>
{products.map((p) => <li key={p.id}>{p.name}</li>)}
</ul>
);
}
Partial Prerendering (PPR)
// app/dashboard/page.tsx
export const experimental_ppr = true;
import { Suspense } from "react";
import { StaticHeader } from "./header";
import { LiveMetrics } from "./metrics";
export default function Page() {
return (
<>
<StaticHeader /> {/* prerendered */}
<Suspense fallback={<MetricsSkeleton />}>
<LiveMetrics /> {/* dynamic, streamed */}
</Suspense>
</>
);
}
ISR with revalidate
// app/posts/[slug]/page.tsx
export const revalidate = 60; // every 60s
export default async function Page({
params,
}: { params: Promise<{ slug: string }> }) {
const { slug } = await params;
const post = await fetch(`https://api.example.com/posts/${slug}`,
{ next: { revalidate: 60, tags: [`post:${slug}`] } }).then((r) => r.json());
return <article>{post.title}</article>;
}
On-demand revalidation
// app/api/revalidate/route.ts
import { revalidateTag } from "next/cache";
export async function POST(req: Request) {
const { tag } = await req.json();
revalidateTag(tag);
return Response.json({ ok: true });
}
Pure React 19 streaming SSR
// server.ts
import { renderToReadableStream } from "react-dom/server";
import App from "./App";
export default async function handler(req: Request): Promise<Response> {
const stream = await renderToReadableStream(<App url={req.url} />, {
bootstrapModules: ["/client.js"],
onError: (err) => console.error(err),
});
await stream.allReady; // remove for true streaming
return new Response(stream, {
headers: { "content-type": "text/html" },
});
}
Client hydration entry
// client.ts
import { hydrateRoot } from "react-dom/client";
import App from "./App";
hydrateRoot(document, <App url={location.href} />);
Cache headers for SSR
// app/api/data/route.ts
export async function GET() {
return Response.json(
{ data: 42 },
{
headers: {
"Cache-Control": "public, s-maxage=60, stale-while-revalidate=300",
},
},
);
}
Server-only utility
// lib/server-only.ts
import "server-only"; // 매 import in client component throws
import { db } from "./db";
export async function loadSecret() {
return db.secret.findFirst();
}
Edge runtime SSR
// app/page.tsx
export const runtime = "edge";
export default async function Page() {
const data = await fetch("https://api.example.com/edge").then((r) => r.json());
return <pre>{JSON.stringify(data)}</pre>;
}
매 결정 기준
| 상황 | Approach |
|---|---|
| Static marketing page | SSG (generateStaticParams) |
| Per-user dynamic | SSR / RSC |
| Mostly static + small dynamic | PPR |
| Data refreshes minutes-scale | ISR with revalidate |
| Internal app, no SEO | CSR (Vite SPA) sufficient |
| Low latency global | Edge runtime SSR |
기본값: Next.js 15 App Router + RSC + Suspense streaming + PPR where applicable.
🔗 Graph
- 부모: Rendering Strategies · Web Performance
- 변형: CSR · SSG · ISR · Streaming SSR
- 응용: Remix · SvelteKit · Nuxt
- Adjacent: React Server Components — 경계 의식 · Hydration · Suspense
🤖 LLM 활용
언제: rendering strategy 결정, SEO + first paint 최적화, RSC + Suspense 설계. 언제 X: 매 internal admin tool with auth-only access (CSR 충분), 매 매우 dynamic real-time app (WebSocket-driven).
❌ 안티패턴
- SSR everything: 매 unnecessary server compute. 매 marketing page → SSG.
- No streaming: 매 await all data → blank for 5s. 매 Suspense + streaming.
- Hydration mismatch: server
Date.now()vs client → 매 warning.suppressHydrationWarning또는 client-only render. - Secret in client component: 매 env var leak.
server-onlyimport. - Massive RSC payload: 매 props 에 huge JSON. 매 boundary 재설계.
- Forgetting cache tags: ISR 인데 invalidation 못 함.
🧪 검증 / 중복
- Verified (Next.js 15 docs, React 19 docs, Vercel blog 2026).
- 신뢰도 A.
🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — SSR full content |